Exploring distributions of TB incidence rates over time by region

dist_tb_inc_region

## Get required packages, managed via pacman
if (!require(pacman)) install.packages("pacman"); library(pacman)
p_load("getTBinR")
p_load("tidyverse")
p_load("viridis")
p_load("hrbrthemes")
p_load("ggridges")
p_load_gh("thomasp85/patchwork", dependencies = TRUE)


tb <- get_tb_burden() 
dict <- get_data_dict()

## Get number of countries with no TB incidence and a global summary
zero_inc <- tb %>% 
  filter(year > 2007) %>% 
  filter(e_inc_100k == 0) %>% 
  group_by(g_whoregion, year) %>% 
  summarise(n = n()) %>% 
  mutate(Region = g_whoregion)

zero_inc <- zero_inc %>% 
  bind_rows(  tb %>% 
                filter(year > 2007) %>% 
                filter(e_inc_100k == 0) %>% 
                group_by(year) %>% 
                summarise(n = n()) %>% 
                mutate(Region = "Global")) %>% 
  mutate(Region = Region %>% 
           factor(levels = c("Global", unique(tb$g_whoregion)))) %>% 
  mutate(type = case_when(Region %in% "Global" ~ "Globally",
                          TRUE ~ "By Region"))

## Plot number of zero TB incidence countries over time
plot_zero_countries <- zero_inc %>% 
  ggplot(aes(x = year, y = n, col = Region)) +
  geom_line(size = 1.1) +
  geom_point(size = 1.5) +
  scale_color_viridis(discrete = TRUE, option = "plasma", direction = -1, end = 0.7) +
  scale_y_continuous(breaks= seq(0, 15, 1), minor_breaks = NULL) +
  theme_minimal() +
  theme(legend.position = "bottom") +
  guides(col = guide_legend(nrow = 2)) +
  facet_wrap(~type, dir = "v") +
  labs(x = "Year",
       y = "Number of countries with zero TB incidence rates",
       caption = "@seabbs Source: World Health Organisation",
       title = "Number of TB Free Countries",
       subtitle = "By region and globally, 2008 to 2016")

## Distribution of TB incidence rates by country stratified by region and year
## Countries with zero incidence rates have been excluded.
plot_nonzero_countries <- tb %>% 
  filter(year > 2007) %>% 
  filter(e_inc_100k > 0) %>% 
  ggplot(aes(y = g_whoregion, x = e_inc_100k, fill = g_whoregion)) +
  geom_density_ridges() +
  scale_x_log10() +
  scale_fill_viridis(discrete = TRUE, option = "plasma", direction = -1, end = 0.9) +
  theme_minimal() +
  theme(legend.position = "none") +
  facet_wrap(~year) +
  labs(x = search_data_dict("e_inc_100k")$definition %>% 
         paste0(" (log scale)"),
       y =  "Region",
       caption = "Countries with zero TB incidence have been excluded.",
       title = "Tuberculosis (TB) Incidence Rates by Region",
       subtitle = "Distribution of country level TB incidence rates, 2008 to 2016")

## Make combined plot
storyboard <- plot_nonzero_countries + 
  plot_zero_countries +
  plot_layout(ncol = 2, widths = c(7, 3))


storyboard

ggsave("storyboard_country_dist_incidnce_rates.png",
       storyboard, width = 16, height = 12, dpi = 330)